Back to All Writing
    Sunday Observer2026-06-28

    Colombo’s gridlock, a scale ceiling

    Originally published in Sunday Observer on 2026-06-28.

    Read Original Article on Sunday Observer

    Core Argument

    Close to two million people move into and through Colombo every weekday, along seven radial corridors converging on a single colonial core, into a Western Province generating around 40 percent of national output on under six percent of the land. It is the engine room of the economy and the place the economy spends its most productive hours sitting still. On 21 April 2026, Lanka Metro Transit put ten low-floor buses on the Makumbura to Pettah and Makumbura to Kadawatha routes, with full passenger operations planned for August. The buses are clean, accessible and digitally tracked, and the public response was loud. The question for a business owner is not whether the buses are good. It is whether the mobility system can scale. This is a productivity constraint with national reach. The Asian Development Bank puts the cost of inadequate transport infrastructure at around 3.5 percent of GDP a year. Average peak speed in central Colombo has fallen to roughly 12 kilometres an hour, slower than a bicycle. The delivery record is the uncomfortable part: Sri Lanka announces transport projects well and executes them at scale poorly. Vehicles arrive. Systems do not. Read through a scale lens, congestion behaves differently at each level. For a micro trader it is inconvenience. For an SME of thirty it is recruitment friction, late deliveries and a tax on every working hour. For a larger firm it caps how many people can be concentrated in one location before the commute sheds the best talent to employers nearer home. Nationally it suppresses the agglomeration effect that makes cities productive. Colombo's density compounds cost rather than output. Singapore is the precedent, and it is not about money. In 2016 it changed the architecture rather than the vehicles. The State took ownership of buses, depots and fare systems, planned routes centrally through a single authority, and tendered operations to firms competing to run the network rather than to own slices of it. Singapore did not out-buy its congestion. It out-organised it.

    What I'd Revise Now

    The column made a testable prediction without framing it as one, and the test has now run. Full passenger operations were planned for August. They did not begin in August. The launch moved to September, and in the interim a further 104 low-floor buses arrived in the country, to be deployed when the project launches. That is the thesis of this piece happening in real time, and faster than I expected. The fleet grew by an order of magnitude while the operating date slipped. Vehicles arrived. The system did not. I wrote that Sri Lanka announces transport projects well and executes them at scale poorly, and the two months since produced a larger fleet and a later start. What I would sharpen is the advice. In June I framed corridor positioning and shift staggering as prudent preparation for a network arriving shortly. On the current record a business should assume the network is a multi-year proposition and plan its location and logistics on that basis, not on the launch calendar. Decentralising ahead of a service that arrives in September is a timing bet. Decentralising because the institution that would run an integrated network does not yet exist is a structural decision, and it is the right one. The buses were never the project. The single planning authority is, and nothing in the past two months has moved it.

    Key Takeaways

    • The Western Province produces around 40 percent of national output on under 6 percent of the land, and spends its most productive hours stationary
    • Inadequate transport costs roughly 3.5 percent of GDP annually, with central Colombo peak speeds near 12 km/h
    • Congestion caps how many people a firm can concentrate in one place before commuting sheds its best talent
    • Buying vehicles without dedicated lanes, fare integration and a single planning authority adds rolling stock to a broken network
    • Singapore's 2016 Bus Contracting Model changed the institution, not the fleet, and that is the transferable lesson
    Visit Original Outlet (Sunday Observer)

    Related Insights